Will AI Save the Economy — or Is It a Bubble About to Burst?

Opinion:

AI firms are funding each other, giving away compute credits, and building billion-dollar data empires. It looks like progress — but is it sustainable?

Artificial intelligence has become the new economic gospel — the idea that smart machines will rescue growth, ignite productivity, and lead us into a new age of abundance.

But look closer, and the story starts to rhyme with another era I witnessed firsthand — the 1990s in Silicon Valley, when computer companies like Compaq were giving away servers and equipment just to seed the market. Back then, the dream was that every startup would need a data center. Today, the dream is that every company will need an AI model.

The technology has changed, but the psychology feels very familiar.

The Dream: AI as the Great Economic Engine

There’s good reason for optimism. For decades, the global economy has been stuck in a slow-growth rut. Productivity — the engine of long-term prosperity — has barely moved. Many economists now see AI as the breakthrough that could finally restart that engine.

1. AI as the new productivity multiplier.
AI doesn’t just replace repetitive work — it amplifies human ability. It can write, reason, design, and analyze at a speed that reshapes industries. Healthcare, logistics, and finance are already feeling its impact, and that’s just the beginning.

2. Investment as stimulus.
The massive investment wave in AI infrastructure — chips, data centers, power grids — acts as a kind of shadow stimulus program. Every new AI cluster built means jobs, materials, and energy demand. Even if the profits take time to appear, the spending itself is fueling growth.

3. Innovation through interconnected investment.
A striking twist this time is that AI companies are investing in other AI companies. The major players — from chip giants to model developers — are buying stakes in each other’s startups. NVIDIA funds AI platforms that, in turn, buy NVIDIA chips. Microsoft invests in AI labs that use Microsoft’s cloud. It’s a closed ecosystem that accelerates progress but also blurs the line between growth and self-dealing.

The Echo: Silicon Valley’s 1990s Deja Vu

I’ve seen a version of this movie before.

In the 1990s, when the internet boom was just starting, computer companies like Compaq were handing out servers to small firms and startups. They weren’t doing it out of charity — they were creating demand. The more servers in circulation, the more people would build on them, and the more the whole market would rise.

That same mentality is back — only now it’s GPUs instead of servers, and data centers instead of web farms. AI giants are effectively underwriting their own customers to keep the ecosystem expanding. It’s brilliant in the short term, but history suggests it can’t last forever.

The Reality Check: When Growth Feeds on Itself

There’s a danger when too much growth is self-referential — when the money in the system keeps bouncing between the same few players.

1. The circular economy of hype.
AI firms fund one another, generating huge paper valuations and press coverage, but much of the spending never reaches real consumers or the physical economy. It’s capital looping around itself.

2. Fragile fundamentals.
If end-user demand doesn’t catch up — if AI tools don’t start driving measurable productivity in factories, hospitals, schools, and offices — the cycle could snap just as it did in 2000. The same infrastructure that now powers the AI boom could become overbuilt and underused.

3. Energy and chip limits.
The physical costs of AI are staggering. Massive energy consumption, global chip shortages, and cooling demands all introduce real-world limits that the hype tends to ignore.

The Inevitable Cycle: Boom, Bust, Rebirth

Technological revolutions always begin as bubbles. The internet did. So did railroads, radio, and electrification.

The overinvestment phase isn’t all bad — it lays the groundwork for future stability. The data centers being built today will outlast the hype cycles. The companies with real business models will survive the shakeout, and they’ll inherit the infrastructure left behind by the frenzy.

AI’s crash, when it comes, won’t be an ending. It will be the start of a more sustainable second act.

The Verdict

AI may not “save” the economy in the way its boosters claim. But it’s undeniably shaping it — pulling capital, talent, and imagination into a new frontier.

If history is any guide, this cycle will end the way all transformative ones do: with a painful correction, followed by quiet reinvention.

And just like in the 1990s, when Compaq was giving away servers to build the internet’s backbone, today’s giveaways — of compute power, cloud credits, and capital — are building the foundations of something real.

The question is whether we’ll learn from history this time… or repeat it.

Opinion Disclaimer:

This article is intended for informational purposes only. The views expressed are solely those of the author and do not constitute financial, investment, or professional advice. Readers are encouraged to consult with their own advisors before making any decisions based on this content.